Re-engineering Social Encounters
Robert Kraut is the Herbert A. Simon Professor of Human Computer Interaction and Social Psychology at Carnegie Mellon University. He received his PhD in social psychology from Yale University in 1973. He conducts research on the design and impact of computer-mediated communication systems. More information is available at www.cs.cmu.edu/~kraut.
I am a social psychologist with a joint appointment in the school of computer science and the business school at Carnegie Mellon University, where I teach and conduct research on the design of computer systems to support human communication and the impact these systems have on the individuals, groups, and organizations that use them. I started my career as a conventional experimental social psychologist. The PhD program at Yale provided rigorous research training and immersion in both classical theory and the cognitive revolution, which was then sweeping social psychology. It instilled in me a standard for empirical research by which I still judge my own work.
Although I found experimental social psychology intellectually exciting, for me, it was the excitement associated with chess – rigorous and demanding, but having limited validity outside the game itself. The goals of the social psychology game, as I understood them then, were to conduct rigorous research, based on discipline-sanctioned methods, in order to answer research questions, whose importance was determined by a consensus of other social psychologists. I was uncomfortable with this self-referential value system, and described my concerns in an essay titled “Social Psychology as Science Fiction,” which I wrote at the end of graduate school. My advisor, Bob Abelson, however, wisely recommended I the bury piece, and nothing but the title remains.
I’ve been interested in applied problems since graduate school, where my first papers were on the development of political activism and deviant behavior. Most of the psychological research questions I thought important seemed to required knowledge and methods from other disciplines as well. I have been fortunate that my academic appointments have been all been interdisciplinary – the sociology department at the University of Pennsylvania, a joint appointment in psychology and sociology at Cornell, and my current position at Carnegie Mellon University. Interdisciplinary programs in academia expand the knowledge base and tools available to answer research questions. At Carnegie Mellon, for example, I have been able to collaborate with computer scientists, who bring new tools for measuring human behavior over long time periods (Kraut, Scherlis, Mukhopadhyay, Manning, & Kiesler, 1996). I have also collaborated with economists, who contribute econometric techniques for analyzing time-series data (e.g., Kraut, Mukhopadhyay, Szczypula, Kiesler, & Scherlis, 1999) and formal modeling as a way to express theory (e.g., Kraut et al., 2002).
In addition to these academic appointments, I have also worked as a researcher in industry, at Bell Laboratories and Bell Communications Research (Bellcore). Industrial research shifts one’s focus from a discipline (or even multiple disciplines) to problems. Indeed, at Bellcore, experimental and social psychologists, sociologists, linguists, anthropologists, communication scholars, and computer scientists collectively comprised the weekly, brown-bag “psychology lunch.” In this environment, being a psychologist was less a matter of pedigree and more a matter of being concerned with designing technology based on human needs.
Perhaps the most exciting aspect of working in an interdisciplinary environment with computer scientists and engineers is that one need not take for granted the environmental constraints under which most social behavior occurs. By adopting an engineering mindset that one can decompose and reassemble reality, researchers are encouraged to rethink aspects of reality that psychology often takes for granted. For example, psychologists have long documented the influence of physical proximity between individuals on social outcomes ranging from the efficiency of referential communication in dyads (Clark & Brennan, 1991), to the likelihood of friendships (Newcomb, 1961), to success of group work (Allen, 1966). In most of this research, physical space is treated an undifferentiated whole. The challenge of designing, socio-technical systems, such as communication systems for distributed groups performing physical tasks, has challenged me to better understand which features of physical space and the face-to-face interaction it enables are essential to successful group work and which are epiphenomena (Kraut, Fussell, & Siegel, in press). I have also been able to work with engineers, computer scientists and other psychologists to develop communication systems to support group work at a distance (Fish, Kraut, Root, & Rice, 1993).
Allen, T. (1966). Managing the flow of scientific and technological information. Cambridge, MA: Massachusetts Institute of Technology.
Clark, H., & Brennan, S. (1991). Grounding in communication. In L. Resnick & J. Levine & S. Teasley (Eds.), Perspectives on socially shared cognition (pp. 127-149). Washington DC: American Psychological Association.
Fish, R. S., Kraut, R. E., Root, R. W., & Rice, R. (1993). Evaluating video as a technology for informal communication. Communications of the ACM, 36(1), 48-61.
Kraut, R. E., Fussell, S. R., & Siegel, J. (in press). Visual information as a conversational resource in collaborative physical tasks. Human-Computer Interaction.
Kraut, R. E., Mukhopadhyay, T., Szczypula, J., Kiesler, S., & Scherlis, W. (1999). Communication and Information: Alternative Uses of the Internet in Households. Information Systems Research, 10(4), 287-303.
Kraut, R. E., Scherlis, W., Mukhopadhyay, T., Manning, J., & Kiesler, S. (1996). HomeNet: A field trial of residential internet services. Communications of the ACM.
Kraut, R. E., Sunder, S., Morris, J., Telang, R., Filer, D., & Cronin, M. (2002). Markets for Attention: Will Postage for Email Help? Proceedings, Conference on Computer Supported Cooperative Work. New York: Association of Computing Machinery.
Newcomb, T. (1961). The acquaintance process. New York: Holt, Rinehart, & Winston.
A Vision of Interdisciplinary Research
Michael Scheessele is an assistant professor in the computer and information sciences department at Indiana University – South Bend. He received a PhD in quantitative and mathematical psychology from Purdue University, a master’s in quantitative and mathematical psychology from Purdue, and a master’s in computer science from DePaul University.
At the start of the 2002-03 academic year, I became chair of the cognitive science committee at Indiana University – South Bend. This is one of those ‘service’ committee jobs we all do, and as someone who loves cognitive science, I was happy to take my turn this year. As a second-year Assistant Professor in the Department of Computer and Information Sciences, it dawned on me that I had come full circle from where I had started nine years earlier. In late summer of 1993, I worked full-time as a senior software engineer for IDX Systems Corporation and was also trying to wrap up an MS in computer science at DePaul University in Chicago. I had enrolled in DePaul’s master’s program in order to improve my job situation, as well as for the intellectual stimulation that it provided. I wound up getting hooked on artificial intelligence, which led me to a course in cognitive science, offered by the CS department in the summer of 1993. During one especially interesting lecture, I decided cognitive science was for me – realizing that I was most interested in that artificial intelligence which is psychologically and biologically plausible. So by fall of the following year, I had completed the MS in CS, quit my job, sold my car, given away or stored many of my possessions, and was enrolled in the quantitative and mathematical psychology PhD program at Purdue University (today my workstation is named ‘quixote’ in honor of fools who mistake windmills for giants).
The transition from a computer science environment to a Psychology environment was not overly traumatic for several reasons:
- Quantitative and mathematical psychology tends to attract people with non-psychology backgrounds.
- My PhD advisor at Purdue, Zygmunt Pizlo, holds both a PhD in electronics from the University of Warsaw and a PhD in psychology from the University of Maryland. (His advice: “Don’t settle for just being the best psychologist in a roomful of computer scientists and the best computer scientist in a roomful of psychologists – try to be the best psychologist in a roomful of psychologists and the best computer scientist in a roomful of computer scientists.”)
- My master’s and PhD committees each included one computer scientist.
The transition from psychology back to computer science has not been difficult either. First, my research is highly interdisciplinary. This research is in the area of vision science, a sub-field of cognitive science. Work in vision science exemplifies the best spirit of interdisciplinary research. For example, AI/computer vision researchers tend to be well-versed in the contributions of Gestalt psychology, while perceptual psychologists often use techniques, like edge detection, that were developed by the AI/computer vision community. In fact, I have used ‘pyramid’ techniques, commonly found in the computer vision literature, to build a computational model of human perception of partially occluded figures. Second, there are two other computer science professors here working on artificial intelligence problems, and this sometimes naturally leads to discussions involving psychology. Both professors serve with me on the cognitive science committee and this in turn provides additional opportunity to interact with others on campus, from biology, philosophy, physics, and psychology, who are interested in the workings of and connections between mind, brain, and machine.
I occasionally have lunch with various professors in the psychology department. When I first set up my research, they gave me direction on where to obtain equipment and permission to use their labs for my psychophysical experiments. In turn, I offered one of the professors my software for running signal detection experiments. As with my computer science colleagues, I find the psychology people on campus to be a great bunch and hope to collaborate with them in the future.
As for the drawbacks and benefits of being a psychologist in a computer science department, one drawback is due to having been out of mainstream computer science for awhile. Since knowledge turns over quickly in this field, I have had to do a fair amount of ‘catch up’. There are also subtle differences between cultures of which one should be aware. For example, computer science people (everywhere) tend to favor publishing in conference proceedings rather than in journals. Some computer science and engineering conferences are as prestigious and selective as journals, but have shorter turnaround time from submission to publication (this allows more papers). As for benefits, I feel that my training as a psychologist enhances my ability to teach computer science courses, because I have insight into how humans process information. (Another benefit of working in a computer science department is that our budget for buying toys seems to be larger than that for other departments!) In closing, I find my position in a computer science department to be flexible enough to allow me to function as both computer scientist and psychologist. Thus, I am quite happy to have come full circle.
Stover Snook has been a Lecturer at the Harvard School of Public Health, Department of Environmental Health, for almost 30 years. At the same time, he was employed as a researcher for the Liberty Mutual Insurance Company. His research centered on occupational health and safety, and more specifically, musculoskeletal disorders. His PhD is from Tufts University.
When I received my bachelor’s degree from Hartwick College in 1954, the Korean War had ended, but the Cold War continued. I was drafted within weeks of graduation and although newly married, I was assigned to the 7th Army in Germany. My new father-in-law, an employee of Bell Telephone Laboratories, sent me a copy of the company bulletin, containing an article that described the new field of human engineering, a.k.a. engineering psychology.
While still in Germany, I wrote to three companies mentioned in the article, and eventually was hired by Dunlap and Associates, a small Connecticut consulting firm with military and space contracts.
With support from the company and the Korean Bill, I received my master’s in experimental psychology from Fordham University in 1960. I remained at Dunlap for six years, participating in the human factors design of the Atlas, Polaris, Hawk, and Sparrow missile systems, the control systems for the Strategic Air Command and the Tactical Air Command, the cockpit of the F4H aircraft, and the Weather Observing and Forecasting System. It was an exciting job. I was on the leading edge of technology, but I became disenchanted with the work on weapons systems. I would rather save lives than terminate them.
My search for a position that would use my skills in a commercial, non-military setting led me to the Liberty Mutual Insurance Company. Liberty Mutual was the largest writer of workers’ compensation insurance in the United States. The company maintained a small research center in Hopkinton, Massachusetts, dedicated to helping people live safer, more secure lives.
I stayed at the research center for 35 years, conducting ergonomics research in areas that were most important to Liberty Mutual and its customers. I was able to conduct experiments on low back pain, manual handling tasks, repetitive motion tasks, heat stress, fatigue, machine guarding, stairway design, and personal protective equipment. Since ergonomics is an interdisciplinary field of study, my research team included a physiologist, an engineer, and a physician assistant/physical therapist. Unlike many commercial companies, Liberty Mutual permitted publication of all results in scientific journals. Liberty Mutual also supported my PhD degree in experimental psychology, received from Tufts University in 1969.
In 1974, I received a part-time appointment at the Harvard School of Public Health. It was an ideal situation – one foot in industry, one foot in academia. My students were primarily physicians, including many flight surgeons from the military, but I also taught nurses, physical and occupational therapists, and industrial hygienists from all over the world.
Teaching health care providers can be an interesting experience. New data often conflict with established opinions. I had to remind my students that only 15 percent of all medical interventions are based on hard, scientific data. The other 85 percent is not necessarily wrong, but it is unproven. For example, in spite of many opinions, the cause of low back pain is unknown. Most treatments in routine use for low back pain are ineffective and some are worse than no treatment at all. Many health care providers can find that message difficult to accept.
I developed a semester course on ergonomics/human factors, and a short course on industrial ergonomics. The semester course was required for many of the graduate programs in occupational health and safety, and the short course was the outreach program to the industrial community. It was a wonderful opportunity to bounce ideas and hypotheses off students, and incorporate their contributions in the research design.
I found it rewarding to see some of my students attain high-ranking positions in the Department of Health and Human Services – and several of my international students returned home to leadership positions in their own countries. Not all of my students came from Harvard. I was able to participate as visiting faculty in numerous universities from Bologna to Taiwan.
At Liberty Mutual, my research made extensive use of psychophysical measures in the experimental investigation of work-related pain, fatigue, discomfort, and overexertion. These are variables with few objective measures. For example, I used psychophysics to develop maximum acceptable weights and forces for lifting, lowering, pushing, pulling, and carrying tasks. A similar methodology was used to determine maximum acceptable forces for repetitive motions of the hand and wrist (flexion, extension, ulnar deviation, and hand grip). A study of stairways used psychophysics to determine the optimal dimensions of risers and treads. At Harvard, I used psychophysics to probe different locations on the face for sensitivity to pressure, in an effort to determine the optimal design of respirators.
I continuously collaborated with company engineers in the application of research results to industrial settings.
Overcoming established beliefs was also an issue with engineers. One of my missions was to collect good, objective data in the laboratory, and then use them to support new ideas in the workplace. For example, good workplace design is more effective in reducing low back disorders than training the individual how to lift. The acceptance of new ideas is much easier when the supporting data are good.
Our research team also conducted epidemiology and cost studies, and evaluations of prevention and treatment. Our studies were among the first to determine the incidence and average cost of compensable low back pain and upper extremity cumulative trauma disorders. Other studies investigated the increased fluid levels in the intervertebral disc during the early morning, and the increased vulnerability to low back pain at that time. A randomized, controlled trial of reducing lumbar flexion during the early morning resulted in significant decreases in the intensity and duration of low back pain.
In 1997, I retired from the Liberty Mutual Insurance Company as assistant vice president and director of the ergonomics laboratory.
It was a very fulfilling journey to conduct practical research in the laboratory, and then discuss and interpret the results in the classroom. However, the journey is not quite over – I continue to lecture at the Harvard School of Public Health.
Food for Thought: A Career in Sensory Evaluation and Consumer Research
Howard Schutz is an emeritus professor in the food science and technology department at the University of California, Davis. His research focused on consumer psychology, in the areas of consumer behavior, food product development, and consumer research methods. His PhD is from the Illinois Institute of Technology.
After a BA in Psychology from the University of Illinois, as an experimental and physiological PhD student at Illinois Institute of Technology in Chicago (who was running out of my GI bill), I was motivated to take a job at the Quartermaster Food and Container Institute (QMF&CI) Acceptance Branch (1951) conducting research in taste and olfaction. The work evolved into research in the area of the measurement of food attitudes and launched me on a career in consumer psychology.
Since the Food Acceptance Branch had several social and experimental psychologists, I was able to build a very good team working relationship which was valuable as a guideline throughout my career. The other divisions of the QMF&CI were staffed by chemists, engineers and food technologists, so I had my first opportunity to blend my psychological knowledge with other disciplines.
After receiving my PhD, I was lured by the emerging field of human factors and moved on to Battelle Memorial Institute in Columbus, Ohio (1957), in the Systems Engineering Division working with other psychologists and engineers. The engineers were a bit leery of psychologists but at least willing to see what they might have to offer, especially in the area of visual displays and other topics where the human factor could play an important role. With my background in the food acceptance area, I wound up working on cooperative projects in the Biosciences Division and eventually moved to that area and developed a sensory evaluation facility. Here I interacted with chemists and biologists and contributed to grant writing in areas where the measurement of characteristics of food by people, including food acceptance, played a significant role.
After five years of the frustrating grant process, I was attracted to the idea of either an academic or industrial position. After interviewing at both UC Davis’ Food Science and Technology Department and at Hunt Wesson Foods Research and Development Department, I decided the excitement (and money!) of industry was too attractive to pass up.
At Hunt-Wesson (1962) I held several positions, at first setting up a psychologically sophisticated sensory evaluation function, then as an Associate Director of Research and finally as a Manager of New Product Exploration and Screening. In these positions I had the opportunity of working with food technologists, engineers, marketing and marketing research people. Over time I was able to recruit other psychologists to work in the sensory lab and with them develop a number of innovative measurement techniques.
With a change in top management from a more quantitative to qualitative approach, it was clear that for me it was no longer an appropriate venue. The fork in my career road was either to look for another industry position or again consider the academic life. With contacts I had built up with UC Davis, it was possible to obtain a professorship in the Consumer Sciences Department (College of Agricultural and Environmental Sciences) (1970) to build courses and a research program in consumer behavior. Thus, I began a 23-year tenure developing courses in consumer behavior, food product development, and consumer research methods. While contact with the psychology department was minimal, Iworked with a number of scientists from other disciplines including home economics, nutrition, textiles and clothing, food science, agricultural economics, sociology, and political science. I also had the opportunity to collaborate with an applied psychologist from the psychology department in connection with the activities of a Center for Consumer Research where we used both qualitative and quantitative methods to study a number of current consumer issues.
I took an early retirement in 1993, but continued in part-time positions at UC Davis, first as director of a Science and Society undergraduate teaching program and then in my current job as a special assistant to the dean of University Extension. In this latter position I have been responsible for assisting in the development of part-time masters degree programs and a Web-based certificate program in sensory evaluation. For the last 10 summers I have been working under contract to the consumer research group at the U.S. Army Natick R&D lab in Natick, MA. Ironically, this is the successor facility to my first job at the QMF&CI. Again I am working in happy concert with a group of psychologists.